CrowdStrike Builds Falcon Guardian Security Into Anthropic's Claude Cowork

The cybersecurity company made the announcement at its Fal.Con Europe 2026 conference in Barcelona. It said the two companies developed the integration together in response to demand from enterprise customers.

How The Integration Works

According to CrowdStrike's statement, Falcon protection will cover Claude agents across their whole lifecycle, from creation through to the tasks they carry out. Oversight runs through the Falcon platform's security control plane and works alongside the controls already in Cowork, including its virtual machine, permissions and administrator settings.

CrowdStrike said the system is designed to spot malicious activity and possible data exfiltration as agents run tasks, handle sensitive information and connect to other enterprise systems. Companies can deploy Cowork with Falcon protection already in place on Windows and Mac, so security teams do not have to fit it after agents have gone into use.

The two companies split the work. Anthropic enabled Falcon to be deployed inside the Cowork environment, and CrowdStrike engineered and tested the integration, covering installation, real-time protection and ongoing updates, the statement said.

What The Companies Said

"The next wave of enterprise AI adoption will be driven by agents," said Daniel Bernard, CrowdStrike's chief business officer. He said customers want the security they already trust to extend to the places where agents operate.

Paul Smith, Anthropic's chief commercial officer, said: "Companies move faster with AI when their security teams can see what it's doing and control how it's used." He added that CrowdStrike customers will be able to run Falcon inside Cowork.

The release also quoted Ben Brophy, chief information security officer at Mondelēz International, who said that as AI use spreads, "security will be essential to unlocking its full value."

CrowdStrike's statement includes a standard caveat that features not yet released are still in development and may change, and that customers should base purchasing decisions on features that are currently available. It did not give pricing.

Why It Matters

As companies move from chat assistants to agents that can open files, run code and act inside business applications, chief information security officers are under pressure to show they can see and control that activity. Products like Cowork give agents direct access to company data and systems, and a compromised or misused agent can do far more damage than a chatbot.

That risk was in the news on the same day. Binance's report on its $6.3 billion hack, published on November 3, said attackers used an internal AI tool to take over the systems of the exchange's wallet engineers.

The deal also shows Anthropic leaning on established security vendors to win large enterprise contracts. The company is preparing for a stock market listing and is telling investors about its fast-growing revenue. Putting a widely used endpoint security tool inside Cowork removes one obstacle for companies that already run CrowdStrike and have been cautious about letting agents work inside their networks.